Welcome
Public Policy · Data Science · Machine Learning
Assistant Director at Singapore's Ministry of Health. Previously Data Scientist at LinkedIn, and Senior Manager for Data Strategy at the Smart Nation and Digital Government Office. Stanford MS, Cambridge BA.
Welcome to my personal website, where you can find my portfolio and publications, and where I experiment with side projects and apps.
My work sits at the intersection of public policy, data science, and machine learning. I'm interested in how technology and data can be applied to improve lives at scale — from national AI strategy to eldercare service design.
Reach me at joseph.lee@cs.stanford.edu.
Lead two teams at the Ageing Planning Office: the Sector Planning team (3 staff) looking at strategy for reshaping the eldercare sector to help seniors age well; and the Centre-based Care and Social Support team (5 staff) overseeing Active Ageing Centres and Senior Care Centres service models.
Led research on how jobs are changing and the future of skills, culminating in the publishing of the interactive data article Future of Skills.
Developed the Public-Private Data Sharing Strategy, the National AI Strategy, and the Public Sector Data Security Review Committee Report. Drove Singapore's data initiatives including the TRUST Healthcare Platform and Apple-HPB LumiHealth project. Developed tech policies for COVID-19 such as TraceTogether.
Led research on how jobs are changing and the future of skills at LinkedIn's Economic Graph Research and Insights team.
Lead author of NLP paper on neural text style transfer by adapting methods from grammar and language correction. Published June 2019.
Co-author of economics networks paper studying how bounded rationality affects social networks through a controlled experiment.
Tutorials explaining deep learning concepts in a simple, intuitive way — published on Medium.
AI-powered push-up and sit-up counter for Singapore's IPPT fitness test. Uses TensorFlow.js MoveNet pose estimation to track reps in real-time via your camera, with IPPT scoring and form feedback.
Full-stack Machine Learning Developer for this grammar and style checker, part of Andrew Ng's Stanford Machine Learning Group. Eventually spun out into the startup Sapling.
Consultant for a large bank to predict Net Promoter Score (NPS) from customer service calls by analysing sentiment within voice calls.
Worked with BJ Fogg on how behavior works and how to design for specific behaviors; provided consultation for companies looking to use Behaviour Design.
Class project combining NLP with representation learning on graphs (GraphSAGE) to predict Amazon item categories.
Class project with Stanford Medical School using Computer Vision to identify and explain whether an abdominal X-ray shows abnormalities.
Class project working with the US Air Force, featured in the Wall Street Journal.
I also build live apps — check out what I'm currently working on.
View Apps →Focus on Machine Learning (researcher under Andrew Ng's Stanford Machine Learning Group) and entrepreneurship. Involved in two start-ups: iDoctor and Crio (now Sapling).
Adam Smith Prize for top student and Adam Smith Dissertation Prize for the top dissertation. President of TEDxCambridgeUniversity.